Chapter 5: UV Photodetection in ZnO Nanowires
5.2. Effect of Rapid Thermal Annealing
generation rate and oxygen re–adsorption rate become constant resulting in a steady photocurrent. It is known that adsorption process is slower than the photodesorption process. Therefore, during UV illumination, not all the holes recombine with the electrons present in the ionized oxygen. As a result, excess holes are available for recombination with the exciton related free electrons. During photocurrent decay, the exciton related electron–hole recombination dominates, which corresponds to the faster decay component, so the photocurrent initially decreases very rapidly. With the surface re–adsorption of oxygen, the photocurrent comes to the initial dark current value very slowly.
-12 -8 -4 0 4 8 12
-20 -10 0 10 20
1.4 1.6 1.8 2.0 2.2
Dark Current (nA)
Bias Voltage (Volt)
As-grown RTA700C RTA800C
I~V
1.0
I~V
3.0
(a) As-grown (b)
RTA700oC RTA800oC Linear fit 1 Linear fit 2
Log(Current)
Log(Voltage) VT
Figure 5.5: (a) The dark current–voltage characteristics of as–grown and RTA treated ZnO nanowires processed at 700° and 800°C. (b) Plot of this I–V characteristics of selected region in log–log scale.
dependence of current on voltage is expected. This behavior is consistent with previous reports on doped metal oxide thin film based resistive switching devices238,239 and it is explained on the basis of presence of oxygen vacancy related traps within the bandgap of the material. Under a negative bias voltage, oxygen vacancies with positive charges migrate away from the interface between contact and the ZnO, which widens the depletion layer, resulting in high resistivity. On the other hand, with positive bias voltage, the oxygen vacancies start moving toward the interface, resulting in higher current. Therefore, the presences of native surface defects (oxygen vacancies) could be identified from the dark I–V characteristic. In case of RTA treated NWs, a nearly linear dark I–V characteristic with slightly high current is observed. Therefore, the analysis of the dark I–V characteristics of the ZnO NWs clearly indicates the presence of high density of trap centres on the surface of the as–grown ZnO NWs, while the trap density is significantly reduced in RTA processed ZnO NWs.189
5.2.2. Photocurrent Spectra
Figure 5.6(a) shows the PC spectra of the RTA treated ZnO NWs processed at 700°C, measured at a bias of 2.5V. After RTA at 700°C, the PC at 369 nm reaches a maximum value of 82.1 µA from that of 9.6 µA for the as–grown NWs and full width at half maxima (FWHM) is reduced compared to the case of as–grown NWs. As a consequence, the photosensitivity is increased to ~24.2×103, leading to an enhancement of a factor of five. With RTA at 800°C, a similar high PC is obtained (~ 84.1 µA) with photosensitivity of ~24.2×103 [Fig. 5.6(b)]. Note that the enhancement in
photosensitivity is significant in this case and is similar to the earlier reports, as shown in Table 1.3 in Chapter–1, where more complex structures were fabricated by surface passivation of the ZnO NWs with different inorganic/organic materials. In our case, observed enhancement factor is quite high compared to the previously reported value for the conventional furnace annealed ZnO NWs.155 Here we obtained a high photosensitivity by employing a simple RTA process, which shows the effectiveness of this process. In the PL spectra of the RTA processed NWs, we observed a reduction in oxygen related defect states after RTA treatment. Therefore, RTA induced reduction in defect density is primarily responsible for the enhanced photoconduction, because very
0 25 50 75 100
300 400 500 600 700
0 25 50 75
0.0 0.4 0.8 1.2 1.6
10k 20k 30k 40k
Experimental Gaussian Fit Component
Photocurrent (A)
(a)
(b)
Wavelength (nm)
(c)
Photosensitivity
Light Intensity (mW/cm2)
Figure 5.6: The photocurrent spectra of the RTA treated ZnO nanowires processed at: (a) 700°C, and (b) 800°C, respectively at 2.5 V bias. (c) The variation of photosensitivity at different intensity of UV light (wavelength 369 nm) for the ZnO nanowires RTA treated at 800°C.
few photogenerated carriers can be trapped inside the defect states and most of the excess carriers contribute to the PC. It is observed that the trap centres related PC peaks in the visible region are considerably reduced for the RTA treated NWs. Figure 5.6(c) shows the UV light (wavelength 369 nm) intensity dependence of photosensitivity of the ZnO NWs, RTA treated at 800°C. The photosensitivity plot shows a sub–linear behavior with the UV light intensity. The observed sub–linear behavior is because of the complex interplay of electron–hole generation, carrier trapping, and recombination within the semiconductor NWs. The high photosensitivity even in low light intensity is an indication of very low value of detection limit. Therefore, we obtained a high value of photosensitivity without making any complex structures, which is advantageous for the fabrication of UV photodetectors.
5.2.3. Photoresponse
To study further the effect of RTA on the PC time response, we performed the photoresponse measurement of the RTA treated NWs under similar conditions. The obtained results of the RTA treated NWs are shown in Fig. 5.7. The photoresponse of the ZnO NWs shows a bi–exponential growth and bi–exponential decay behaviors. The growth and decay time constants are calculated from the bi–exponential fitting to the experimental data and the obtained parameters are summarized in Table 5.1. The time dependent growth behavior of the photoresponse curve is fitted with the equation,
1 ⁄ ⁄ ……… (5.1)
where I1, A1 and A2 are positive constants. The first exponential term corresponds to the electron–hole generation process and the last exponential term represents the oxygen adsorption process. Calculated time constants from fittings are τ1 =24.7s and τ2 =295.0s for the as–grown NWs indicating a very rapid photocurrent growth initially followed by a very slow decay process. For the RTA treated NWs, the photocurrent growth and decay time constants are 11.4s and 261.8s, respectively for 700°C annealing and 12.3s and 107.0, respectively for 800°C annealing. In this process, the electron–hole generation is the only source for current carriers and all other processes decrease the carriers.
Similarly, when the UV light is turned off, the photocurrent show a rapid decay followed by a slow decay. The time dependent decay behavior can be fitted with a bi–exponential decay equation,
0 300 600 900 1200 15001800 0
20 40 60 80
0 900 1800 2700 3600 0
20 40 60 80
0 900 1800 2700 3600 0
3 6 9
0 900 1800 2700 3600 0
20 40 60 80
(d)
Photocurrent ()
Time (sec) (c)
Photocurrent ()
Time (sec)
UV ON UV OFF
(a)
Exp. data Bi-exponential Fit
Photocurrent ()
Time (sec)
UV ON UV OFF
(b)
Photocurrent ()
Time (sec)
UV ON UV OFF
Figure 5.7: The photocurrent growth and decay behaviors of the ZnO nanowires: (a) as–
grown, (b) RTA processed at700°C, (c) RTA processed at 800°C, respectively. Photocurrent growth and decay are measured under the illumination of 360 nm UV light. (d) Photocurrent growth and decay of the RTA treated ZnO nanowires (700°C) under the pulsed (300 s) UV light illumination at a bias voltage of 2.5 V.
∞ ⁄ ⁄ ...……… (5.2)
where A3 and A4 are positive constants and ∞ refers to the photocurrent after infinitely long time of the decay experiment, which essentially is the dark current. The first exponential term in eqs. 5.2 corresponds to the electron–hole recombination process Table 5.1: Summary of the growth and decay time constants of the ZnO nanowires, obtained from the bi–exponential fitting to the experimental data points.
ZnO NWs
PC Growth time constants (s)
PC Decay time constants (s)
τ1 τ2 τ3 τ4
As–grown 24.7±0.4 295.0±0.8 25.7±0.5 347.9±1.1 RTA700°C 11.4±0.3 261.8±1.1 12.3±0.3 298.5±1.1 RTA800°C 12.3±0.5 107.0±1.1 13.6±0.3 118.4±0.5
and the last exponential term represents the oxygen adsorption process. The decay time constants are calculated to be 25.7s and 347.9s for the as–grown NWs, 12.3s and 298.5s for the RTA treated at 700°C, 13.6s and 118.4s for the RTA treated at 800°C, respectively. Therefore, the PC growth as well as decay becomes faster after the RTA treatment. The calculations of individual time constants show that electron–hole recombination as well as generation rates become double after RTA at 700°C and do not change significantly with higher temperature annealing. On the other hand, oxygen adsorption rates during the photocurrent growth as well as decay are systematically decreased. Similar bi–exponential decay behavior with time constants of several seconds has been reported by several groups for the ZnO nanobelts, NWs and thin film.18,106,186,191,240 Figure 5.7(d) shows the photoresponse of the RTA treated NWs under periodic UV illumination. It is observed that the maximum photocurrent in the next cycle is slightly increased compared to the previous cycle because of the incomplete growth and decay of the PC during the measurement cycle. Second and third cycle of photocurrent growth and decay show exactly the replica of first cycle, indicating a reproducible PC response of the RTA treated ZnO NWs, which is important for the real time application in photodetectors.
As the RTA processing significantly reduces the surface defect related trap centres and modified the surface of the ZnO NWs, the band bending is less here compared to the as–
grown NWs case resulting in a comparatively higher conductivity, as revealed in the dark I–V characteristics. As a consequence, the photocurrent reached the saturation value very fast. Therefore, the structural improvement caused faster photocurrent growth and decay from the RTA treated ZnO NWs. This is consistent with the PL results discussed earlier.